Purification and Properties of Rat Liver Microsomal Stearyl Coenzyme A Desaturase (fatty acid desaturation/membrane-bound enzymes)

Transcription

1 Proc. Nat. Acad. Sci. USA Vol. 71, No. 11, pp , November 1974 Purification and Properties of Rat Liver Microsomal Stearyl Coenzyme A Desaturase (fatty acid desaturation/membrane-bound enzymes) P. STRITTMATTER, L. SPATZ, D. CORCORAN, M. J. ROGERS, B. SETLOW, AND R. REDLINE Department of Biochemistry, University of Connecticut Health Center, Farmington, Conn. 632 Communicated by Eugene P. Kennedy, August 29, 1974 AB STRACT The terminal enzyme of the NADHdependent stearyl coenzyme A desaturase system has been isolated from rat liver microsomes. This desaturase is a single polypeptide of 53, daltons containing 62% nonpolar amino-acid residues and one atom of non-heme iron. The purified protein forms high molecular weight aggregates that can be dispersed by detergent procedures. Desaturase activity requires NADH, stearyl coenzyme A, oxygen, lipid, and the three enzymes, cytochrome b5 reductase (EC ), cytochrome b5, and desaturase. Cytoehrome b5 is the direct electron donor to the desaturase, which appears to utilize the iron in the oxidationreduction sequence during desaturation of stearyl coenzyme A. The characterization of the liver microsomal stearyl coenzyme A desaturase system by workers in several laboratories (1-11), using spectral methods and partially purified components from hen and rat liver microsomes has shown that the oxygen- and NADH-dependent formation of oleyl coenzyme A involves cytochrome b5 reductase (EC ; NADH : ferricytochrome b5 oxidoreductase), cytochrome b5, and a terminal, cyanidesensitive oxidase, the desaturase. The present report describes the isolation of the terminal component, the desaturase, from liver of rats induced for this enzyme. A sequence of extractions of contaminating proteins from microsomes, followed by solubilization of the desaturase by detergents and subsequent purification, results in a protein preparation that appears to be homogeneous by several criteria. It contains one molecule of non-heme iron, which appears to participate in stearyl coenzyme A desaturation in the presence of oxygen and reduced cytochrome b5. METHODS AND MATERIALS Sodium deoxycholate (Matheson, Coleman and Bell) was purified as described (12). Sephadex G-75 was obtained from Pharmacia, and Whatman DEAE-cellulose (DE-52), from Reeve Angel. Triton X-1 was the purified product of Packard Instrument Co. The ['4C]deoxycholate was obtained from Mallinckrodt Chemical Works, the NADH, stearyl coenzyme A, and bathophenanthroline sulfonate were products of Sigma Chemical Co., and [14C]stearyl coenzyme A was a product of New England Nuclear. The egg lecithin was obtained from Supelco in a chloroform solution. We prepared clear liposomes by removing solvent in a nitrogen stream, adding.2 M Tris-acetate buffer, ph 8.1 (Tris buffer), to yield a suspension of 12.5 mg/ml, sonicating under nitrogen at for 15-2 min at 4 mw, and centrifuging the slightly turbid suspension for 15 min at 1, X g to yield the clear liposome suspension, which was stored under nitrogen at and used for no longer than 1 days. Abbreviation: Tris buffer, Tris-acetate buffer (ph 8.1) Cytochrome b5 was prepared from either calf or rabbit liver microsomes by the detergent extraction procedure that yields the complete form of this heme protein (13). The proteins from the two species were indistinguishable as electron donors for the desaturase. The soluble catalytic fragment of cytochrome b5 was prepared by tryptic digestion of cytochrome b5 (13). Either rabbit or calf liver cytochrome b5 reductase was prepared iby the procedure originally described for isolation from ra liver (12). The preparation of the catalytic fragment of the reductase has also been described (14). The method of Weber and Osborn (15) was used for acrylamide gel electrophoresis in 9% crosslinked gels and in the presence of.2% sodium dodecyl sulfate after preincubation for 1 hr in 1% sodium dodecyl sulfate and mercaptoethanol. Protein bands were stained by the procedure of Fairbanks et al. (16). Lipid-extractable phosphorus was measured by the method of Chen (17), protein by the procedure of Lowry et al. (18), and reducing sugar by the phenol-sulfuric acid method (19). Flavin determinations involved measurement of the fluorescence of a trichloroacetic acid extract or a solution of desaturase in 1% sodium dodecyl sulfate at 53 nm with excitation at 47 nm. A slight modification of the bathophenanthroline sulfonate procedure (2) was used for iron analysis. Initial electron spin resonance measurements were made with a Jeol model JES-M spectrometer at 1K. Acid-labile sulfide was determined as described by King and Morris (21). Amino-acid analysis was preformed by the technique of Spackman et al. (22), and cysteine content was determined as cysteic acid (23). Tryptophan and tyrosine were determined by the spectral procedures of Beaven and Holliday (24) in 1% sodium dodecyl sulfate. NH2-terminal amino-acid analysis used the microdansyl technique after two sequences of subtractive Edman degradation (25). Assay Procedures. The procedure of Jones et al. (4) with 14C-labeled substrate was used to standardize the rapid spectral method previously described for microsomal preparations (26). In the latter case, reoxidation of cvtochrome b5 at 424 nm, after NADH oxidation in the presence and absence of stearyl coenzyme A, was followed, as suggested by the studies of Oshino et al. (7). With purified desaturase and other components, the disappearance of NADH fluorescence with and without substrate also provided a continuous assay and initial rate data. For all assays, the added complexity is the requirement for proper preincubation of the desaturase at various stages of purification with the other protein and lipid components of the system. At fractions 1-5 in the preparation, 5,l of enzyme was brought to 2.3% sodium deoxycholate by the addition of an appropriate volume of 1% sodium deoxycholate in.2

2 4566 Biochemistry: Strittmatter et al. M Tris buffer at 25C. Quickly, 2 Ad of 2,uM cytochrome b5 and 1 um reductase (or the catalytic fragment of reductase) in.2 M Tris buffer and 3 /J of sonicated egg lecithin suspension (12 mg/ml, in.2 M Tris buffer) were added in succession. Within 3-4 sec,.19 ml of.1 M Tris buffer was added to achieve the dilution of detergent required for activity in the subsequent assay. At fractions 6-9 the desaturase was in an appropriate detergent mixture. The order of addition of components at 25C was then 3.ul of desaturase preparation, 2 d1a of the cytochrome and reductase solution, 3,ul of egg lecithin suspension, and.19 ml of the.1 M Tris buffer. The final purification steps (fractions 1 and 11) were carried out either in the presence or absence of egg lecithin. When no phospholipid was added, 2-4 IAI of the purified enzyme was mixed with.5,ul of 1% Triton X-1 and incubated for 1 min at C, 2 M1 of cytochrome and reductase solution were then added at C and incubation was continued for another 1 min before addition of 3,Al of egg lecithin. After an additional 2 min at C,.19 ml of Tris buffer at room temperature was added and the assay procedure was initiated. When egg lecithin was added during the final purification steps,.7,ul 1% sodium deoxycholate in.2 M Tris buffer was added to 2,ul of the cytochrome and reductase solution and incubated for 1 min at C. The purified desaturase (4,u, containing phospholipid) was then added and again incubated at C for 1 min before addition of.19 ml of the Tris buffer. Cytochrome b5 reoxidation at 424 nm at 25C was used routinely (26). The blank rate of NADH oxidation was first determined by addition of 1 ul of 1.5 mm NADH in.2 M Tris buffer to any one of the preincubated assay mixtures. The cytochrome b5 was reduced rapidly and then reoxidized after exhaustion of the NADH. The initial linear rate of cytochrome reoxidation was extrapolated to obtain the time for complete reoxidation of the NADH and reduced cytochrome b5. The NADH addition was then repeated after the addition of 3,ul of 2 mm stearyl coenzyme A to the assay cuvette, and the time for complete reoxidation was again measured. The difference between the time for reoxidation in the presence and absence of substrate was then used to calculate nmol of oleyl coenzyme A formation per min, assuming a stoichiometry of 1 mol of NADH per mol of oleyl coenzyme A formed (26). The blank NADH oxidation, usually between 2 and 3% of the rate with substrate, includes reductase and cvtochrome b5 autooxidation as well as an appreciable cyanidesensitive autooxidation, presumably of the desaturase (see Results). The other two assay procedures were conducted in a similar manner except that ['4C]oleate was measured in one case, and NADH fluorescence disappearance in the other. The agreement among the three methods was 4±1%. Purification of Desaturase. Fifty Long-Evans strain rats, weighing between 125 and 15 g, were starved for 48 hr, fed Purina laboratory chow for 24 hr, starved a second 48-hr period, and refed with Nutritional Biochemical Corp. "Fat Free" test diet for 2 hr on a schedule that permits the animals to be killed at the beginning of a day. All procedures were carried out at -5C with centrifugation at 12, X g unless otherwise specified. Livers were removed quickly from 5 rats (36-42 g) during a 2- to 3-min period and placed in cold.25 M sucrose, 1 mm Trisacetate, 1 mm EDTA buffer, ph 8.1. The liver was washed twice with 1 volumes of the sucrose buffer, blotted, weighed, Proc. Nat. Acad. Sci. USA 71 (1974) minced with scissors, and homogenized with 6 ml of sucrose buffer per g of liver in glass homogenizers with loose-fitting Teflon pestles. Mitochondria, nuclei, and cell debris were removed as a pellet by centrifugation at 18, X g for 15 min. The supernatant fluid was diluted with 3 volumes of distilled water, and 1 ml of.8 M CaCl2 were added per liter of preparation. Centrifugation of this suspension for 5 min at 5 X g served to pack the microsomal cell fraction. The clear supernatant fluid was discarded, and the microsomal pellets were suspended in sufficient.1 M Tris buffer to yield 2 ml of suspension after homogenization. After addition of 8 ml of.5 M EDTA, ph 8.1, the suspension was centrifuged for 3 min. The supernatant fluid was discarded and the pellets were homogenized with sufficient.2 M Tris buffer to yield a volume of microsomal suspension half the original weight of liver (fraction 1). This suspension was brought to.45% sodium deoxycholate by the addition of 1% sodium deoxycholate in.2 M Tris buffer, sonicated for 4 sec at 6 mw, and centrifuged for 6 min. The supernatant fluid was carefully removed with a syringe and discarded. The relatively loosely packed material above the glycogen pellet was then removed with 6 ml of.2 M Tris buffer by agitating with a Vortex mixer. The suspension (about 1 ml) was brought to.75% sodium deoxycholate by further addition of the 1% deoxycholate solution (assuming that the wet pellets were at.45% deoxycholate), and sonicated for 4 sec (fraction 2). After fraction 2 was centrifuged for 6 min, the supernatant fluid was again removed carefully and the yellow-brown layer was resuspended with 36 ml of.2 M Tris buffer to yield about 55 ml of suspension, which was brought to 1% deoxycholate by addition of the 1% deoxycholate solution and sonicated for 4 sec (fraction 3). The loosely packed pellet obtained by centrifuging this suspension for 6 min was suspended with 36 ml of.2 M Tris buffer and sonicated for 4 sec to yield about 5 ml of fraction 4, which was again centrifuged for 6 min. The yellow-green pellet was suspended in 1 ml of.2 M Tris buffer, sonicated for 4 sec, and stored overnight (fraction 5, about 16 ml). Before fraction 5 the enzyme preparation cannot be stored overnight without loss of 2-5% of activity. Two volumes of 2.5% Triton X-1 were added to fraction 5 and after sonication for 4 sec the suspension was allowed to stand for 3 min before centrifugation for 6 min. The pellet was suspended with 9 ml of 2.5% Triton X-1 and sonicated for 4 sec. Slowly, first.2 ml of 1% sodium deoxycholate in.2 M Tris buffer and then.19 ml of.8 M CaCl2 was added with stirring (fraction 6). After 3 min, this suspension was centrifuged for 3 min. The supernatant fluid (about 9 ml) was brought to 25% propylene glycol concentration by the addition of the concentrated solvent and again centrifuged for 3 min to yield about 13 ml of supernatant fluid (fraction 7). Fraction 7 was layered on a 5 X 2-cm DEAE-cellulose column equilibrated with.1 M Tris buffer and eluted with the same buffer. The desaturase emerged in the void volume, and 13 ml were collected and centrifuged for 3 min to yield 13 ml of supernatant fluid (fraction 8). This enzyme solution was either stored at -7 C or immediately placed on a 15 X 2-cm Sephadex G-75 column that was equilibrated with.2% sodium deoxycholate,.1 M Tris buffer. The column was developed with the same buffer, at a flow rate of about 9 ml/hr. The protein emerged in the void volume ( ml,

3 Proc. Nat. Acad. Sci. USA 71 (1974) Stearyl Coenzyme A Desaturase 4567 TABLE 1. Purification of desaturase (a) (b) Specific activity Protein Activityt (nmol/mg Fraction* (mg) (nmol/min) per min) , , , , , , , , Not done Not done , ,1 355 * Fractions 1-1 were from one preparation; fraction 11 was from another preparation. t By cytochrome b5 reoxidation (see Methods and Materials). fraction 9), and the Triton, propylene glycol, and most of the residual lipid were included and did not begin to elute until ml. Low phospholipid enzyme was isolated by diluting fraction 9 with 2 volumes of.1 M Tris buffer and collecting the precipitate after 15 min by centrifugation for 6 min. The precipitate was washed free of much of the deoxycholate by resuspending with 15-2 ml of.1 M Tris buffer, homogenizing in an all-glass homogenizer, and centrifuging for 3 min. This precipitate was homogenized with 2-3 ml of.1 M Tris buffer and stored at C or -7'C (fraction 1). Alternatively, the enzyme was isolated with phospholipid by adding 1.3 ml of the egg lecithin suspension (16 mg of phospholipid) to fraction 9 before dilution with the.1 1\I Tris buffer. The same centrifugation and washing procedure as described for the egg lecithin-free preparation was then used to obtain fraction 11, which was also stored at C or -7"C. RESULTS The purification procedure (Table 1) results in a 12- to 13- fold purification with a 3-4% yield at fraction 8. Subsequent manipulation of the desaturase, to remove Triton and some tightly bound phospholipid, does not result in further apparent purification and involves appreciable loss of protein. The observed activities must be considered minimal values, since the major problem in assaying the enzyme as purification proceeds is proper dispersion of the highly aggregated desaturase. Disc-gel electrophoresis (Fig. la) shows the course of purification and provides some indirect data concerning the disaggregation-activity relationship. It is clear that one major polypeptide is concentrated during purification. By use of appropriate protein standards, this polypeptide, under denaturing and reducing conditions, has an apparent molecular weight of 53,. Therefore, in the absence of high concentrations of detergents, e.g., fraction 1, the native polypeptide must form high-molecular-weight aggregates. Fig. lb shows one type of evidence that identifies this polypeptide with the desaturase. In this experiment rats were divided during the feeding period into those that were starved but not refed the 4p 1: I --" " -,- IN. '1 -. VW - 'Oklilo- W.4& 1T FIG. 1. Acrylamide disc-gel electrophoresis (a) of fractions 1-8 and 1 of a typical preparation with two protein concentrations of fraction 1; and (b) of fraction 7 obtained from starved (-) and starved-refed (+) rats. See Methods and Materials for details and text for protocol for (b). "fat-free" diet (-) and those that were refed (+). Liver from both groups was subjected to purification to fraction 7, and only in the induced animals (+) does a concentrated polypeptide appear at the position of the major component of fraction 1. There is some indication in Fig. la that fraction 1 may, in fact, be slightly less contaminated with minor, particularly smaller, peptide bands than fraction 8. Thus, some purification may occur on the Sephadex column which is not seen as an increase in specific activity because the preincubation conditions may fail to properly disperse all of the aggregated fraction 1 to yield desaturase which can interact rapidly with reduced cytochrome b5. This assumption is consistent with the observation that fractions 1 and 11 from several preparations varied in specific activity from 24 to 36. Preparations of fraction 1 contained 18.4 ± 1.8 nmol of iron/mg of protein. This corresponds to a minimum molecular weight of 55, +E1%. Consistently, the preparations with slightly lower iron content also had lower specific activities. Iron loss appears to result from desaturase instability during the 2-hr Sephadex gel filtration procedure. Fraction 1 also still contains 2-1 mol of phospholipid/mol of iron. However, there was no detectable acid-labile sulfur, fiavin, heme, or reducing sugar (<.2 mol per mol of protein) in these preparations. Preparations of fraction 11, to which phospholipid was added, contained about mg of phospholipid/mg of protein. By including [14C]deoxycholate in the latter steps in the purification during the isolation of desaturase when phospholipid is re-added, we found that 3 mol of the detergent per mol of enzyme appeared in fraction 11 despite the washing procedure used. The spectrum of the purified desaturase at fraction 9 is shown in Fig. 2. Only a major protein band with a high tryptophan content, indicated by the 29-nm shoulder, is seen. The low absorbance in the 3- to 43-nm region is similar to that of ribonucleotide reductase (27), which also contains non-heme iron. The low extinction at nm is consistent with the absence of heme. Since radioactive Triton X-1 was not available to us, the exact content of this detergent is difficult to ascertain. Nevertheless, the fact that Triton X-1 is included in the Sephadex column (see Methods and Materials) and only begins to appear more than 2 ml after the protein ib am

4 4568 Biochemistry: Strittmatter et al. Proc. Nat. Acad. Sci. USA 71 (1974) (i c um < Wavelength, nm FIG. 2. Spectrum of fraction 9 at C. has been eluted, and the absence of detectable absorbance characteristic of the Triton X-1 absorption spectrum in deoxycholate (maximal at 28 and 275 nm), suggest that little Triton remains in fractions 1 and 11. The single major absorption band is accounted for by the spectral contributions of 24 tyrosyl and 17 tryptophanyl residues, as determined by the spectral methods of Beaven and Holliday (24) and by tyrosine analysis (see below). The amino-acid analysis of fraction 1 (Table 2) emphasizes the predominance of nonpolar residues in the desaturase. By the procedure of Capaldi and Vanderkooi (27), 62% of the amino acid residues are nonpolar. This provides an explanation for the marked tendency of the protein to polymerize in aqueous solvents during isolation. NH2-terminal end group determinations by two cycles of Edman degradation on preparations of fractions 1 and 11 show a dominant NH2- terminal serine followed by tryptophan. Low concentrations of several other residues (<1%) were detected, but these appear to arise from small peptide contaminants, since prior gel filtration on Bio-gel P-1 equilibrated with 1% sodium dodecyl sulfate removes them completely. Then only serine residues followed by tryptophan are seen by the dansyl technique as the NH2-terminal sequence. The 53,-dalton polypeptide seen on disc-gel electrophoresis, therefore, appears to be a single polypeptide species. The essential components for NADH-dependent stearyl coenzyme A desaturase are the three enzymes (reductase, cytochrome bn, and desaturase), the substrates (NADH, oxygen, and stearyl coenzyme A), and phospholipid. Detergents appear to serve the function of dispersing the proteins, in particular, the desaturase. If phospholipid was omitted in the assay of fraction 1, in which Triton X-1 is the detergent (see Methods and Materials), 3-35% of TABLE 2. Amino-acid composition* of fraction 1 Lys 23 Glu 29 Met 12 His 14 Pro 21 Ile 21 Arg 22 Gly 3 Leu 58 Asp 33 Ala 35 Tyr 24 Thr 2 Cys 8 Phe 28 Ser 33 Val 25 Try 17 Total residues = 458 Molecular weight = 53,162 * See Methods and Materials for procedures. The Lys residue value was chosen to agree with the minimum molecular weight determined by iron analysis and from disc-gel electrophoresis. Several preparations of fractions 1 and 11 yielded essentially identical results. Composition is given in residues per 23 lysine residues.._, 6S Fe removed (%/6) FIG. 3. Correlation between iron removal and loss of desaturase activity. A suspension of fraction 11 (12.7 nmol of iron), 16 nmol of cytochrome b5, and 2 nmol of reductase in 2.65 ml of.1 M Tris buffer was incubated for 3 min at 32 C. After the mixture was cooled to 25 C, 2,1 of 1 mm bathophenanthroline sulfonate was added to 5-Al aliquots of the enzyme solution. To five such samples,, 15, 3, 45, and 6 nmol of NADH were added, respectively, and incubation at 25 C was continued until the NADH was oxidized, as indicated by cytochrome reoxidation. Aliquots of each sample were then assayed for activity and the amount of bathophenanthroline iron couplex formed. maximal activity was observed. Whether this resulted from the low residual lipid in fraction 1 or represents activity of the enzyme system in this detergent alone, is not yet clear. We were able to confirm the previous report of Shimakata et al. (11) that there is no detectable enzyme activity when the soluble catalytic fragment of cytochrome b5 is substituted for the complete amphipathic heme protein. In contrast to Shimakata et al. (11), however, we find that in the absence of cytochrome b5 neither reductase nor its catalytic fragment can serve as direct electron donor, although either form of the reductase can be used to reduce the cytochrome b5 rapidly in the normal assay system. A catalytic role for the non-heme iron of the desaturase is implied by the observed loss of activity upon iron removal by a ferrous chelator (Fig. 3). It was found that incubation of a preparation of fraction 11, which contained about 1.5 mg of lipid per mg of protein and 3 mol of deoxycholate per mol of desaturase, with reductase, cytochrome, and desaturase at 32 C for 3 min was sufficient to obtain 7-8% of the desaturase activity. When such activated complexes of the three proteins and phospholipid were preincubated with a relatively high concentration of ferrous chelator, iron was removed only if the incubation mixture contained NADH. Iron in fraction 11 appears to be reduced by the NADH, reductase, and cytochrome b5. The high concentrations of bathophenanthrolene sulfonate and desaturase during this incubation, before a 2-fold dilution for activity assay, results in iron removal proportional to the total amount of NADH added and, thus, the duration of desaturase reduction. Both the dependence of iron removal upon reduction by substrate and the direct correlation with loss of catalytic activity suggests a catalytic function for the non-heme iron in the desaturase. Moreover, as in the case of the enzyme in microsomes (3, 7), the activities of the purified enzyme fractions are completely and reversibly inhibited by.5 mm KCN. Preliminary electron paramagnetic resonance studies at 1 K with 7 nm desaturase as well as 2 nm enzyme reduced enzymatically with NADH as in Fig. 3, showed no iron signals. However, these data must be considered inconclusive, since both higher protein concentrations with oxidized and reduced

5 Proc. Nat. Acad. Sci. USA 71 (1974) enzyme and lower temperatures may be necessary to detect a catalytically active high spin iron intermediate. DISCUSSION The present studies with purified stearyl coenzyme A desaturase confirm earlier reports on microsomes and the partially purified enzyme (1-11) that reduced cytochrome b5 is the normal electron donor for the desaturase in endoplasmic reticulum. The requirement for phospholipid reported by Jones et al. (4) is also observed with the purified enzyme, even in the presence of Triton X-1. Both the tendency of this protein to form high-molecular-weight, inactive aggregates during purification and the instability of the enzyme, particularly in detergents, account for the difficulties encountered in the development of the isolation procedure. These properties of the enzyme required rapid isolation and storage at -7'C in low detergent concentrations. Dispersion of aggregates of the desaturase for assay was a balance between detergent conditions that would disperse aggregates completely and still avoid denaturation. As in the case of the other two components of the desaturase system, the desaturase appears to be a single polypeptide. However, it is larger (456 residues), and the high content of nonpolar residues in the complete protein (62%) contrasts with the flavo- and heme proteins. The latter are amphipathic proteins each containing a hydrophilic, catalytic segment and a smaller hydrophobic sequence (12, 13) involved in attachment to microsomal vesicles. Since the dominant polypeptide in the final desaturase preparations (fractions 1 and 11) appears to be nearly homogeneous by several independent physical and chemical criteria and to contain functional iron, it is unlikely that a minor protein contaminant is responsible for the desaturase activity. The maximum activities observed with such preparations yield a turnover at 25C of 21 mol of substrate per mol of desaturase per min. It is not yet clear, however, whether or not this represents a true maximum velocity in this complex system. The observed turnover number and the total activity in induced rat liver microsomes indicates that the desaturase may represent.7-.8% of the total microsomal protein. The dietary induction of desaturase described by Oshino and Sato (8) and used here, thus results in at least a 2-fold increase in the concentration of this protein to form a significant quantitative component of the endoplasmic reticulum. The apparent involvement of iron in the catalytic activity of the desaturase is remarkably similar to the situation reported recently by Pistorius and Axelrod (28) for soybean lipoxygenase as well as other non-heme iron proteins (29, 3). Here, too, the enzyme contains 1 mol of iron, which is removed under reducing conditions by orthophenanthroline. These authors also detected an electron paramagnetic resonance signal at g = 6 upon addition of linoleic acid to the lipoxygenase in the presence of oxygen. The utilization of the purified desaturase to probe mechanistic questions regarding interactions with cytochrome bs, oxygen, and substrates must include similar electron paramagnetic resonance experiments Stearyl Coenzyme A Desaturase 4569 under optimal conditions. Isolation of the desaturase also provides the third component required to reconstitute a complete liposomal desaturase system for a detailed examination of the essential protein and lipid interactions in the mobile nonpolar phase. We thank Mr. John Bunch for the careful amino-acid analysis. This investigation was supported by Research Grant GM from the United States Public Health Service. 1. Holloway, P. W., Peluffo, R. & Wakil, S. J. (1963) Biochem. Biophys. Res. Commun. 12, Gellhorn, A. & Benjamin, W. (1964) Biochim. Biophys. Acta 84, Oshino, N., Imai, Y. & Sato, R. (1966) Biochim. Biophys. Acta 128, Jones, P. D., Holloway, P. W., Peluffo, R.. & Wakil, S. J. (1969) J. Biol. Chem. 244, Gurr, M. I. & Robinson, M. P. (197) Eur. J. Biochem. 15, Holloway, P. W. & Wakil, S. J. (197) J. Biol. Chem. 245, Oshino, N., Imai, Y. & Sato, RI. (1971) J. Biochem. (Tokyo) 6, Oshino, N. & Sato, R. (1971) J. Biochem. (Tokyo) 69, Holloway, P. W. (1972) Biochemistry 11, Holloway, P. W. & Katz, J. T. (1972) Biochemistry 11, Shimakata, T., Mihara, K. & Sato, 1R. (1972) J. Biochem. (Tokyo) 72, Spatz, L. & Strittmatter, P. (1972) J. Biol. Chem. 248, Spatz, L. & Strittmatter, P. (1971) Proc. Nat. Acad. Sci. USA 68, Strittmatter, P. (1971) J. Biol. Chem. 246, Weber, K. & Osborn, M. (1969) J. Biol. Chem. 244, Fairbanks, G., Steck, T. L. & Wallack, D. F. H. (1971) Biochemistry 1, Chen, P. S., Toribana, T. & Huber, W. (1956) Anal. Chem. 28, Lowry,. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J. (1951) J. Biol. Chem. 13, Dubois, M., Gilles, K. A., Hamilton, J. K., Rebers, P. A. & Smith, F. (1956) Anal. Chem. 28, Atkin, C. L., Thelander, L. & Reichard, P. (1973) J. Biol. Chem. 248, King, T. E. & Morris, R.. (1967) in Methods in Enzymology, eds. Estabrook, R. W. & Pullman, M. E. (Academic Press, New York), Vol. X, pp Spackman, D. H., Moore, S. & Stein, W. H. (1958) Anal. Chem. 3, Ozols, J. (197) J. Biol. Chem. 245, Beaven, G. H. & Holliday, E. R. (1952) Advan. Protein Chem. 7, Weiner, A. M., Platt, T. & Weber, K. (1972) J. Biol. Chem. 247, Strittmatter, P., Rogers, M. J. & Spatz, L. (1972) J. Biol. Chem. 247, Capaldi, R. A. & Vanderkooi, G. (1972) Proc. ANat. Acad. Sci. USA 69, Pistorius, E. K. & Axelrod, B. (1974) J. Biol. Chem. 249, Ruettinger, R. T., Olson, S. T., Boyer, R. F. & Coon, M. J. (1974) Biochem. Biophys. Res. Commun. 57, Fisher, D. B., Kirkwood, R. & Kaufman, S. (1972) J. Biol. Chem. 247,